Research In Image Digital Watermarking Based On Discrete Wavelet Transform

With the development and mature of digital technology, more and more digital works to spread through the Internet, they can be low-cost, high-speed to be copied and disseminated, but these features can also be easily used by the pirates, making network information security risks which endanger all-round way the society's economic, political, cultural and other fields. Therefore, how to protect digital copyrights in the network and the information security of digital works has become an imminent task. And digital watermarking technology is considered to resolve copyright issues an effective way: it make full use of human visual and auditory resolution limitations, embedded with copyright-related information to confirm the copyright of digital products.At present, the carrier of the image that is image watermarking are being the most widely studied in the digital watermarking research fields, and in image compression standard "JPEG₂000", the wavelet is used as a major technology. Therefore, the main contents of this paper is based on image watermarking in wavelet domain, mainly making use of discrete wavelet transform (DWT) theory to image watermarking embedding and detection research, and also discusses and validates the hidden image watermarking and robustness. Specifically, this paper complete the following tasks: First of all, the fundamentals, such as the basic characteristics of digital watermarking, theory and research at home and abroad are analyzed. The theoretical model of digital watermarking is given for the proposed watermarking algorithm, the realization and test,which provides a theoretical basis.Secondly, as the digital watermarking algorithm theoretical basis, the development of wavelet analysis, wavelet transform theory, multi-resolution analysis and Mallat algorithm are introduced in detail; especially focusing on the two-dimensional discrete wavelet transform and image processing; and one typical wavelet-based image watermarking algorithm is given.Finally, according to a typical watermarking algorithm and the integrated improvements, an improved adaptive robust watermarking algorithm of gray-scale images based on DWT is proposed, which makes watermark images and carrier images integrate after the different scale wavelet transform. Also test results show that the algorithm is quite satisfactory for common image processing operation which has a good robustness. In addition, since a significant proportion of color images in practical applications, this paper also puts forward a color image watermarking algorithm of robustness, which combines human visual system (HVS) with different color components embedded in a different way. The experiment result shows that the use of the watermark embedding algorithm has better robustness and invisibility.